Cutting IT costs through changing service levels

When IT is under cost pressure, leaders often look to “do more with less:” to become more efficient or more effective through process improvements, to cut open positions, to reduce training dollars, to renegotiate contracts, or to use short-term tactics like reducing buffer stocks or hiring temps.

These approaches are all what I would call “supply side” approaches. They are changes to how IT as a supplier provides service, and they’re changes that should not be noticed by IT customers and users around campus.

In addition to these supply side approaches, IT needs also to address the demand side. Many times, the administration controlling budgets and asking for cutsĀ does not understand what IT does: they see a $10 million IT operating budget and figure there’s waste in there somewhere.

IT needs to show the services it’s providing to justify its budget. But if costs need to be cut, IT also needs to review its service levels with customers.

Especially in higher education, service levels can be challenging: they can sound like an excuse to provide bad service, or a way for IT to trick its customers. For this reason, sometimes service levels have a “statement of cooperation” in them to acknowledge everyone’s working for the same institution and working for a common purpose.

The thing is, without service levels the default expectation is perfection: 24x7x365 support with 100% capacity hot backups for disaster recovery, immediate ticket resolution, 100% uptime, etc. Without service levels, staff can be blamed (or even fired) if service doesn’t meet these impossible targets. IT staff are naturally risk averse, and therefore services will be designed for perfection.

And guess what? Designing services for perfection is slow, and very very expensive. Networks, servers, databases, applications, and the entire stack can be “over-designed” resulting in huge costs and unneeded capacity.

No matter how much supply side cost reduction you employ, this problem cannot be solved within IT. IT customers and users need to agree before capacity is reduced. But they need to see the benefits, too.

Creating service levels sets expectations for users. Users should know their incidents will be addressed within four hours, rather than “sometime soon.” Users should start to feel a level of certainty that users can plan on.

Creating service levels helps the right services be available at the right time. When users see service levels in context, they may agree that “email is more critical than my departmental event management tool.” They become a prioritization tool.

Creating service levels, especially early in the design of a service, reduces the total cost of ownership for the service and ensures the service isn’t over-designed.

Service levels also help you refocus conversations about cost-cutting. I heard a great presentation by Georgia State at EDUCAUSE 2009 where they were asked to cut their budget and they could say: “which classroom projectors should be removed?” The IT budget becomes less of a target for cost-cutting when people understand how service quality will be affected.

If this post resonates with you, please read also the “Chargeback vs. IT Governance” article for ideas about how to tie costs to services and when to charge for service.